Ad Attribution is crucial to measuring the success of advertising campaigns. With Amazon Marketing Cloud it is possible to change the standard-attribution logic.
Amazon Ads Standard Attribution Models
If you’ve received reports with conversion data from Amazon Ads you’re familiar with attribution models for ASIN and pixel conversions. ASIN conversions use brand-based or merchant-based (seller) models:
- Brand-based is for advertisers using Sponsored Brands, Sponsored Display, Sponsored Products for Vendors, and Amazon DSP.
- Merchant-based is for Sellers using Sponsored Products.
The details below explain attribution, but note that from "custom attribution"-data sources in Amazon Marketing Cloud, it is not possible to generate the same conversion data from the "standard attribution"-data sources.
Standard Attribution: Merchant-Based Ad Products (ASIN conversions)
For Sponsored Products, a last-click attribution model is used, counting the most recent click. The attribution window is 7 days in the Amazon console for Sponsored Products, but accessing the API allows 1, 7, 14, or 30-day windows. Per default, emax digital allows sellers to switch between the standard 7-day window or also use the 14-day window.
Standard Attribution: Brand-Based Ad Products (ASIN conversions)
For brand-based ads (DSP, Sponsored Display, Sponsored Brands, Sponsored Products for Vendors), multiple ads compete for attribution. Ads are eligible if a shopper clicked or viewed them within 14 days before converting. If the shopper engaged with several ads, the model prioritizes the last click over the last view.
Standard Attribution: Pixel Conversions (Amazon DSP)
Pixel conversions are only applicable if there is a valid pixel added to the campaign. Ads tracking a valid pixel are eligible if clicked or viewed within 14 days before conversion. The model uses a last-click approach, prioritizing clicks over views.
Custom Attribution Models with Amazon Marketing Cloud
Custom attribution in AMC allows you to choose your lookback window (up to 28 days) and select the attribution model (see below for details), which determines how conversions are assigned to touchpoints (such as impressions of or clicks on an ad).
Amazon provides standard queries to help compare different attribution models, such as "First Touch", "Last Touch", "Linear", and "Position-Based" models. emax digital can help you choose the model that fits your advertising goals and query the attribution data from the AMC.
Please note: emax digital does not query Custom Attribution per default, but only as scoped custom project.
The table below compares the custom attribution models provided by Amazon as templates in AMC. The models are compared with the following Amazon Ads scenario.
Example Amazon Ads scenario: You reach a cohort of customers through your Streaming TV ad. The cohort of customers that saw your ad go to Amazon.com on the same day and they are exposed to your display ad. One week later, they search for your brand on Amazon, click on your sponsored products ad, and purchase your product.
The path to conversion of this cohort of customers is: "Streaming TV → Display → Sponsored Products → Purchase"
Please note: A common misconception is that custom attribution shows the path to conversion. While ad exposure sequences are used in the query, custom attribution does not provide this insight. To view "path to conversion" you can use the relevant dashboard in emax digital. (The linear and position-based models report weighted sums of attributed conversions by campaign.)
|
Model |
How credit is assigned | Pros | Cons |
| First Touch attribution gives full credit to the first touch point (impression or click) of an ad interaction. | The final output from the instructional query will include one row for each campaign with key metrics such as as the attributed conversions and the conversion rate. Conversions are only attributed to the campaign when the given campaign was the first touchpoint. In the example above, Streaming TV receives 100% of the credit for the purchases. | The first touch model helps you to understand how shoppers are discovering your brand. If your goal is to generate awareness and expand reach, then this attribution model will allow you to see which ad product / campaign is best at attracting shoppers. | First touch doesn’t show the full picture. Other campaigns do not receive any credit. In the multi-touch conversion process, any touch that follows the first touch will not be considered. |
| Last Touch attribution assigns full credit to the last ad a user saw or clicked. | The final output from the instructional query will include one row for each campaign with key metrics such as as the attributed conversions and the conversion rate. Conversions are only attributed to the campaign when the given campaign was the last touchpoint. In the example above, Sponsored Products receives full credit for the purchases. | Last Touch allows you to see which touchpoint was last seen by a shopper before converting. If you are focusing on driving conversions, Last Touch will allow you to see which ad product / campaign is best at converting shoppers. This model may be a good fit for you if you have a short buying cycle. | Last Touch ignores all the touch points before the final interaction. Many of the touchpoints prior to that last interaction will be just as important at attracting relevant shoppers and bringing them down the funnel. |
| Linear attribution assigns each touchpoint in the conversion path with equal credit for the conversion. | The final output from the instructional query will include one row for each campaign with key metrics such as as the attributed conversions and the conversion rate. Conversions are attributed all campaigns with a touchpoint and credit is split evenly. In the example above, there are 3 touchpoints. Each touch receives 33.3% of the credit. | Linear Attribution provides you with a more balanced look at your whole advertising strategy when compared to a single-event attribution model. It helps to demonstrate the value of a multichannel or multi-touch strategy. | Linear attribution assumes all touchpoints are equally important. However, some advertising strategies are more effective than others, and this model will not surface that insight. This model can lead you to overvalue some channels and undervalue others based on the way it assigns credit. |
| Position-based attribution splits credit between the first and the last touchpoints on your customer’s conversion path. Weights assigned to these two touchpoints are customizable. Work with your media / campaign manager to determine the weights for the two touchpoints. | The final output from the instructional query will include one row for each campaign with key metrics such as as the attributed conversions and the conversion rate. Conversions are attributed to the campaign when the given campaign was either the first or the last touchpoint based on weights you assign. The first touch point receives credit (such as 30%) and the last touch point receives credit (such as 70%). For the example above, if you assign 30% credit to the first touch (Streaming TV) and 70% credit to the last touch (Sponsored Products), your measurement values the last touch more than the first touch. | This model captures the impact of top and bottom-of-funnel activities, both of which are critical to the conversion path. | This model assume that the two most important interactions a customer has with your business will be the very first and the very last before conversion, while ignore all the touchpoints in between. |